Quick – Close The Lid!. Phil Wolff outlines the library service [he] want[s]: an Amazon alternative and asks who would build it:
“Version 1
Pass two elements:
- ISBN
- ZIP code
This might look like:
Return:
- The book description and a list of libraries in or near the location that have this book in their collection. Availability info too. Link to each library's page for that book.
Later releases:
- Follow the Amazon SOAP API.
- Extend to non-western languages and non-book holdings.
- Compare prices at local independent stores and on Amazon.
- Pass a report of queries to acquisition desks at local libraries (what's hot that you don't have on the shelf).
In the spirit of ISBN.nu.”
Phil doesn't know it, but he just opened a major pandora's box in libraryland. In a room full of relatives, you don't talk about religion or politics. In a room full of geeks, you don't ask which operating system is best. And in a room full of librarians, you don't really want to ask about Earth's Largest Library.
You see, Amazon, Google, and AskJeeves have all rocked our world, and we're not much further along than we were when Steve Coffman wrote his controversial article in 1999. Which, unfortunately, means the answer to Phil's question is no one. At least, not in the foreseeable future.
It's not totally our fault, though. A little realized fact outside of libraryland is that we're completely beholden to our database vendors. For the most part, we can't build our own software for running library catalogs, mainly because of a lack of money, time, resources, and programmer librarians. So we have to rely on the handful of vendors that make the software that runs our catalogs. Naturally, these vendors don't really play nicely together, and it's a very drawn-out process to switch from one vendor to another so it's relatively rare.
Here's an example: the Virtual Illinois Catalog (VIC). I work on VIC, so if we're ever in a bar, I can tell you the stories-oy-the-stories (for example, how five days before it went live at the 1999 ILA conference where the Secretary of State cut the ribbon, it crashed and wouldn't restart when we tried to add a twelfth catalog). VIC lets you search twelve of the major shared catalogs in Illinois – 600 libraries, more than 40 million items. It uses the Z39.50 protocol to connect to the 12 catalogs which are run by 5 different vendors. Each of those vendors implements what should be a relatively straightforward Z39.50 protocol, and they do it in such strange and magical ways that we can only offer three types of lowest common denominator searches – title, author, and subject. And those are all keyword searches, because some of the vendors don't support phrase searching in Z39.50. We can't let search by ISBN or limit by format.
Our ultimate goal has always been to let you search the holdings of ALL Illinois libraries from within VIC and let you request any items that you find. However, VIC is fundamentally the same today as it was when it went live in 1999. Why? Politics, lack of funding, lack of consensus for how to proceed, lack of standards, and lack of time that the VIC development team can devote to the project. (Side note: VIC is about to be migrated to new software, which should help with some searching and port problems, but won't add the interlibrary loan request functionality, natural language querying, or search limiters we so desperately need).
So if we're having trouble connecting just 13 catalogs for the kind of search Phil describes, you can imagine how cold it will be in h@ll when we're able to connect libraries across the U.S., let alone the world. OCLC's Worldcat is kind of close, but it costs money to participate, and it's not open to the average person. You can only get access through a subscription purchased by your home library, which isn't very Amazon-like in its simplicity. Imagine having to type in a nine-digit authorization number and a password every time you want to search Amazon. As Damon Wayans used to say, “Homey don't play that.”
So if we're dependent on our vendors and they don't want to work together or implement standards in ways that allow for collaboration, and we're too busy being librarians to create our own software, where does that leave us when it comes to Phil's request? Pretty much nowhere. A couple of years ago, the VIC development team contacted Amazon to ask if they'd let us use their software to run VIC, and their answer was no, not even for a price. My guess is that it would take a major (and I mean major) National Science Foundation or National Endowment of the Arts grant to even come up with a plan for something like this. There isn't anybody else stepping up to the plate so unfortunately, Phil, I wouldn't go out and buy that winter parka any time soon. [The Shifted Librarian]